Troubleshooting
Important
All commands given in this chapter have to be run with superuser privileges. Therefore make sure to log in as user root for running the commands.
USB driver issues
If you encounter issues during the USB driver installation, several measures should be taken:
Unplug all USB devices (including the HUB-GM100) except for keyboard and mouse
Uninstall any USB-related drivers and software such as STLink, FTDI drivers and similar – you can also postpone this step to a second iteration
- Start the Windows Device Manager with the
devmgr_show_nonpresent_devices=1environment variable set and click Show hidden devices on the View menu (see Device Manager does not display devices that are not connected for further information) Uninstall any COM ports, STLink devices and other CDC and USB serial devices
Uninstall any USB Composite devices
- Start the Windows Device Manager with the
Remove the registry key
HKLM\SYSTEM\CurrentControlSet\Control\usbflags\1DB60104Restart the computer
Now connect the HUB-GM100 again and check the driver installation process. If you still see an unknown serial gadget device in the Windows Device Manager, double click it, switch to the Driver tab, click Update Driver and choose Browse my computer for driver software. Before continueing, download a special USB driver information file . Back in the driver installation wizard, select the location in which the downloaded file has been saved. Windows should now use the information provided in that file to install the correct driver.
General problems
In case of general problems at the application or operating system level, messages in the system journal should be examined first since they might provide useful information on what went wrong at which time in which service, layer or component.
The first port of call on any Linux system is dmesg. Running this command outputs all log messages issued by the Linux kernel. These messages are usually related to hardware and resources initializations and issues. Any hardware or filesystem errors are reported here.
Since SIINEOS is fully managed by systemd it can be inspected and controlled via systemctl and journalctl. To get an overview of all loaded, active and running units (services, resources, mounts etc.), run systemctl. To list all failed units run systemctl --failed. Running systemctl status gives a summary of the current system state and print a tree view of all running units. To view the status and most recent log messages of one or multiple units, run systemctl status <NAME-PATTERN>, e.g. systemctl status incore-app*.
Additionally all kinds of log messages can be examined through the journalctl command. It can be run without any arguments which outputs both kernel and service messages. To query the messages of a certain unit (service, resource, mount), call it with the -u <UNIT> arguments, e.g.
root@hub-gm:~# journalctl -u incore-apps
-- Logs begin at Thu 2019-02-14 10:12:00 UTC, end at Thu 2020-03-19 00:11:46 UTC. --
Mar 19 00:00:02 hub-gm systemd[1]: Starting Start all InCore apps...
Mar 19 00:00:03 hub-gm sudo[284]: root : TTY=unknown ; PWD=/ ; USER=root ; COMMAND=/usr/bin/systemctl --no-pager -q start incore-app@test
Mar 19 00:00:03 hub-gm sudo[284]: pam_unix(sudo:session): session opened for user root by (uid=0)
Mar 19 00:00:03 hub-gm sudo[284]: pam_unix(sudo:session): session closed for user root
Mar 19 00:00:03 hub-gm systemd[1]: Started Start all InCore apps.
Apps filesystem
A FAT32 filesystem containing the installed apps is mounted at /apps which you can verify by invoking df /apps or mount | grep apps. The output should look like
root@hub-gm:~# df /apps/
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/mmcblk0p1 64366 1218 63148 2% /apps
root@hub-gm:~# mount | grep apps
/dev/mmcblk0p1 on /apps type vfat (ro,noatime,nodiratime,sync,uid=1000,gid=1000,fmask=0022,dmask=0022,codepage=437,iocharset=ascii,shortname=mixed,utf8,errors=remount-ro,discard)
As indicated by the ro flag, the apps filesystem is mounted read-only per default to improve security and reliability and to reduce write cycles to the flash memory. It can be made writable by either activating the InCore development mode or by running
mount /apps -o remount,rw
If the apps filesystem is not mounted properly during the boot process or dmesg indicates filesystem errors, you can try to either repair the filesystem by running
systemctl stop apps.mount
fsck.vfat -a /dev/mmcblk0p1
reboot
If the problem persists after the reboot, the filesystem has to be reinitialized. During this formatting process all data, i.e. all installed InCore apps, are removed and can’t be recovered afterwards. So make sure to back up all required files before. When finished, run the following commands:
systemctl stop apps.mount
mkfs.vfat -n SIINEOS_APP /dev/mmcblk0p1
reboot
After rebooting, the filesystem should be mounted properly and should not contain any data.
Storage filesystem
An F2FS filesystem is mounted at /storage allowing apps to store their data. It also holds all Docker-related data such as images, containers and volumes. To verify that the filesystem is mounted properly and to check the available space, run the df /storage and mount | grep storage commands:
root@hub-gm:~# df -h /storage/
Filesystem Size Used Avail Use% Mounted on
/dev/mmcblk0p6 6.7G 2.2G 4.6G 33% /storage
root@hub-gm:~# mount | grep storage
/dev/mmcblk0p6 on /storage type f2fs (rw,noatime,nodiratime,lazytime,background_gc=on,discard,no_heap,user_xattr,inline_xattr,acl,inline_data,inline_dentry,flush_merge,extent_cache,mode=adaptive,active_logs=6,alloc_mode=reuse,fsync)
The storage filesystem is checked for errors and inconsistencies on each boot. Any problems are fixed automatically whenever possible. If the filesystem is damaged unrecoverably, it has to be reinitialized. During this formatting process all InCore apps and Docker data are removed and can’t be recovered afterwards. So make sure to back up all required files before. When finished, run the following commands:
systemctl stop storage.mount
mkfs.f2fs -f /dev/mmcblk0p6
reboot
After rebooting, the filesystem should be mounted properly and should not contain any data.
Factory reset
Warning
When resetting the device to factory defaults, all installed apps, settings, database files, Docker containers and other custom files will be lost. Please make sure to back up all important files before.
The device can be reset to factory settings by flashing a SIINEOS image to the device (contrary to installing an update bundle only). Please make sure to back up all required files before. When finished, run the following commands to download and flash the device:
cd /tmp
systemctl stop storage.mount
systemctl stop apps.mount
incore-cli download https://download.inhub.de/siineos/images/siineos-standard-armhf-disk-v2.7.7.img.gz
zcat siineos-standard-armhf-disk-v2.7.7.img.gz | dd of=/dev/mmcblk0 bs=4M
sync
Do not interrupt the last command and do not interrupt the power supply of the device while it is running. After the last command has finished, unplug the device from the USB port and any power supplies. Eventually the device should start with a clean SIINEOS 2.7.7 installation.